Bragg peak profiles and diffuse scattering in powder diffraction
- Datum: 12.09.2024
- Uhrzeit: 14:00 - 15:00
- Vortragende(r): Paolo Scardi
- Department of Civil, Environmental & Mechanical Engineering, University of Trento, via Mesiano 77, 38123 Trento, Italy
- Ort: Max Planck Institute for Solid State Research
- Raum: 7D2
- Gastgeber: Dep. Nanochemistry
Powder diffraction has evolved enormously over the last twenty years, both in experimental techniques and in data analysis, but despite the general trend towards increasingly holistic analysis methods, some aspects have not progressed as much. Most analyses, such as Rietveld structural refinement, still employ empirical functions to model the diffraction peak profiles, with little regard for the implications of this choice; and little regard for the missed opportunity to gain additional insights into the microstructure through more in-depth, physically based modelling of the profiles. A further neglected aspect concerns the diffuse scattering component which is often improperly absorbed by the background, invariably treated as a disturbance, again with empirical functions. Yet, it is well known that the intensity "lost" in Bragg scattering due to thermal disorder is transferred to the diffuse component, carrying along physical information on the local lattice dynamic structure, especially relevant in nanocrystalline systems.
In this presentation I show the basics and some examples of an alternative and robust approach to profile modelling, based on physical models of the microstructure with few and stable parameters to be refined (optimized) with the comparison and analysis of experimental data. I also present a diffuse scattering model that captures a basic aspect, namely the correlation between atomic thermal oscillations between the closest atoms. The algorithms and procedures shown are fully compatible with the Rietveld method approach, and in particular with the TOPAS computing platform, among the most used in vast fields of research and industrial technologies.